Tsunamites Versus Tempestites: A Comprehensive Review from the Precambrian to Recent Times

Abstract

Insight regarding the overall geological history of tsunamis and their impacts requires information gained from preserved deposits. Although recent decades have seen a rise in tsunami deposit studies overall, most reviews focus on specific time intervals, such as the Paleozoic, the K–Pg boundary, the Quaternary, or historical and recent events, while others concentrated on particular depositional settings, including lacustrine, offshore, or onshore environments. This review paper provides a comprehensive synthesis of tsunami deposits spanning the geological record from the Precambrian to recent times based on a global compilation of onshore, offshore, and lacustrine examples. Selections from the available evidence is traced from the oldest known tsunamites in the Archaean through major extinction boundaries such as the K–Pg, to the well-preserved Holocene and historical deposits. The findings indicate that while the fundamental sedimentological signatures of tsunamis have remained broadly consistent over geological time, their recognition in ancient strata remains challenging due to difficulty in differentiating between storm deposits (tempestites) and other high-energy facies. A central aspect of this review is the critical assessment of diagnostic criteria proposed to differentiate tsunamites from tempestites. By using a multidisciplinary approach, integrating sedimentological, paleontological, geochemical, and geomorphological evidence in palaeotsunami research, this review provides a detailed framework to improve the confidence in identifying tsunami deposits. This, in turn, enhances palaeotsunami reconstructions, which are valuable for advancing hazard assessment along vulnerable coastlines.

1. Introduction

As a scientific discipline, palaeotsunami research remains relatively young. It is still in the process of refining its methodological frameworks and establishing robust criteria for the identification of tsunami deposits. As the field evolves, there has been an increasing emphasis on integrating multidisciplinary approaches, including sedimentology, palaeontology, geomorphology, geochemistry, geophysics and recent/historical records—to enhance the accuracy of tsunami reconstructions. Seminal works by Atwater [1] and Dawson et al. [2] were pivotal in initiating a fundamental paradigm shift in the understanding of tsunami generation and associated sedimentation processes. This shift was further accelerated by the catastrophic tsunamis of the Indian Ocean (2004), South Pacific (2009), Chile (2010), and Japan (2011) [3,4], which profoundly transformed scientific perspectives on tsunami hazards and underscored the critical need to reassess existing models (e.g., [5,6,7]). In the aftermath of these events, researchers focused on sedimentological evidence, coastal geomorphology, and the integration of geological and historical data in tsunami hazard assessment [8,9,10].

Over the last three decades, researchers have shown a surge of interest in studying tsunami related deposits with case studies spanning worldwide regions [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25]. Consequently, this rapid build-up of different datasets and perspectives motivated updated comprehensive reviews aimed at synthesizing and summarizing the state-of-knowledge [19,26,27,28,29].

Despite the increasing volume of research on tsunami deposits from the Quaternary, historical, and recent times, evidence from older geological periods, particularly from the Precambrian (>541 Ma) to the Neogene (23.03–2.588 Ma), remain scarce and poorly documented [30,31,32]. According to Dawson et al. [26], this gap can be attributed to two primary factors: (i) the low preservation potential of tsunami deposits, as the coastal and shallow marine environments most impacted by such events are frequently subjected to continuous sediment reworking; and (ii) the inherent difficulty in distinguishing sedimentary structures generated by tsunamis from those produced by other high-energy marine processes, such as storms. The latter factor presents a major challenge for researchers in the field. That is why recently, many authors have increasingly focused on the comparative analysis of tsunamites and tempestites across various depositional settings, including onshore [17,20,21,24,25,33,34,35,36,37,38], offshore [11,15,19,39,40,41], and lacustrine systems [42,43,44,45]. These efforts aimed to refine diagnostic criteria and improve the reliability of distinguishing between these two types of high-energy sedimentary deposits. According to Sztanó et al. [43] and Doukani et al. [25], an additional challenge in palaeotsunami research is to accurately determine the triggering mechanism of a tsunami based only on the sedimentological and taphonomic characteristics of its deposits. However, recent and historical tsunami deposits are often directly linked to well-documented triggering events, such as large-magnitude earthquakes, submarine landslides or volcanic eruptions. However, establishing these cause-and-effect relationships becomes much more complex when it comes to older geological records, for which direct observational data are lacking. In these cases, interpreting the origin of tsunami deposits requires careful integration of sedimentological, stratigraphic and tectonic (seismic, volcanic, etc.) information. An additional, though less frequent, triggering mechanism for tsunami events is extraterrestrial impacts, such as the strike of a meteorite. Such events can generate large-scale tsunamis, and this mechanism is often associated with tsunamites preserved near the Cretaceous-Paleogene (K–Pg) boundary [46,47,48,49,50,51,52] (Figure 1).

The aim of this paper is to present a comprehensive review of the current state of knowledge regarding tsunami deposits preserved in onshore, offshore, and lacustrine environments, spanning from the earliest known geological records to recent historical events. Special attention is given to the methodological advances and challenges associated with the identification and interpretation of tsunami deposits, particularly in distinguishing them from other high-energy event deposits, such as those generated by storms (tempestites). By synthesizing recent research findings and addressing key diagnostic criteria, this review seeks to highlight existing gaps, ongoing debates, and emerging perspectives in palaeotsunami studies, with the goal of contributing to a more robust framework for recognizing and interpreting tsunami-generated sedimentary records across diverse geological settings.

2. Methodology

This review adheres to the geological time framework established by the International Commission on Stratigraphy, spanning from the Precambrian through the Holocene [53]. The synthesis is structured chronologically, beginning with the Precambrian and progressing through the Paleozoic and Mesozoic. A dedicated subsection addresses the Cretaceous-Paleogene boundary, given its global significance and the abundance of impact-related tsunami studies. From the Cenozoic onward, a finer subdivision is applied, treating the Paleogene and Neogene together and devoting special emphasis to the Pleistocene and Holocene to recent time, where the volume and quality of evidence are greatest.

The present review is based on a comprehensive and critical literature search, namely published articles, book chapters, and monographs, and also “grey literature” (e.g., unpublished PhD and MSc thesis), selected where they explicitly referred to tsunami deposits (tsunamite, tsunamiite) or where they employed broader terminology related to high-energy or catastrophic events, including homogenites, megaturbidites, megatsunamis, mass flows, seismites, and related concepts. Such a broad set of descriptions was adopted to ensure full coverage of the diverse terminologies used across different sub-disciplines of sedimentology, stratigraphy, and geochemistry.

Bibliographic data were compiled through systematic searches of major scientific databases, including Web of Science, Scopus, and GeoRef, supplemented by publishers’ platforms such as Elsevier, Springer, and MDPI. Additional references were retrieved through Google Scholar and ResearchGate, especially for grey literature and regionally focused works. Search queries were applied across titles, abstracts, keywords, and, where possible, full texts. Following the removal of duplicates and works of limited relevance, a final corpus of literature was assembled. Although not exhaustive, this review is considered representative of the present state of research, recognizing that variable terminology and the incomplete preservation of ancient deposits inevitably limit coverage.

The collected works were analyzed with three main objectives: (i) to document the sedimentological and paleontological evidence used to infer tsunami deposition (e.g., erosional bases, grading patterns, sedimentary structures, mixture of organisms, taphonomy); (ii) to evaluate the interpretive certainty, which was classified qualitatively: cases interpreted exclusively as tsunami deposits are regarded as highly reliable; those where tsunami origin is proposed alongside alternative mechanisms (such as storms or debris flows) are treated as moderately reliable; weakly supported cases are those where attribution to tsunamis remains tentative, studies explicitly rejecting a tsunami interpretation are also considered, as they highlight the methodological challenges in differentiating tsunamites from other high-energy deposits; (iii) to examine the proposed triggers across different ages, including seismic events, submarine landslides, volcanic activity, extraterrestrial impacts, and climate-driven processes.

Whenever possible, localities were consolidated into broader paleogeographic domains to reduce heterogeneity and highlight regional trends. For example, closely spaced sites within the same basin or coastal setting are treated together, even when multiple case studies are available. However, we do not combine events with different ages into a single example. Instead, we present together studies from the same region (e.g., Cabo Verde) or from nearby sites within the same basin or coastal environment, especially when they fall within the same age interval. This allows us to reduce heterogeneity and highlight regional trends without merging distinct events. Similarly, temporal clusters were identified in intervals where tsunami evidence is most abundant, such as at major extinction horizons, the Permian–Triassic boundary (P–T), the Cretaceous–Paleogene boundary (K–Pg), or within well-dated Quaternary sequences.

Ultimately, this review aims not only to compile occurrences of tsunami deposits through time but also to assess their diagnostic criteria, evaluate the certainty of their interpretation, and compare them against alternative high-energy depositional processes. By integrating cases across multiple geological periods and depositional settings, the study provides a global framework to understand how tsunami deposits are recognized, interpreted, and differentiated from other catastrophic sedimentary records.

3. Tsunami Deposits in the Geological Record

Extreme events such as tsunamis have frequently inundated the world’s coastlines [29] and can have a major impact on marine environments [54,55]. According to Schnyder et al. [56], the geological record should reflect a comparable or even higher frequency than that observed in modern and historical times. However, evidence of such deposits is rarely preserved or identified in older strata [26,57]. Despite the challenges of identifying tsunamites in the geological record, due to factors such as post-depositional alteration, limited preservation potential, and the similarity of tsunami deposits to those produced by other high-energy processes, several authors have interpreted some sedimentary sequences as the result of tsunami events across different geological periods [58] (see Figure 2). These interpretations are typically based on a combination of sedimentological and palaeontological criteria, integrated with tectonic and volcanic context, as well as evidence of abrupt marine inundation extending into the onshore environment during run-up, or offshore deposits associated with backwash flows (

Table 1. Listing of tsunami records from the Precambrian to recent times, including age, triggering mechanism, location, and environmental settings. Most of the records are cited and discussed in the main text.

Age	Location	Criteria	Trigger	Environment	Reference
Archaean (3.48 Ga)	Australia	Large imbricated clasts, hummocky bedding, Bouma-type graded sequences.	Unknown	Below storm-wave base	[59]
Archaean (3.2–3.5 Ga)	South Africa	Impact spherules, shock-deformed minerals.	Meteorite impact	Shallow-marine	[60]
Late Archaean (2.6 Ga)	Australia	Anorbital wave ripples.	Meteorite impact	Deep marine shelf	[61]
Palaeoproterozoic (2.5–1.6 Ga)	India	Tidal successions with seismites and tsunami beds; includes deformed cross-bedding, convolute lamination, slump folds, and faults.	Seismic activity	Tidal to shallow-marine basin	[62]
Mesoproterozoic (1.8–1.6 Ga)	China	Rip-up clasts, poorly sorted gravels, fining-upward units, erosional bases.	Seismic activity	Coastal to shallow-marine	[63]
Mesoproterozoic (ca. 1.45 Ga)	USA	Molar-tooth structures, granular lime mud with terrigenous silt, swept ooids and rounded, coarse-grained feldspathic quartz sand, and erosive current with gutter casts.	Seismic activity	Below storm-wave base	[64]
Cambrian (541–485 Ma)	Argentina	Coarse-grained strata, allochthonous input, conglomerate interbedded within burrowed mudstones, and deep scour.	Faulting or rifting beyond the shelf margin	Continental shelves and epeiric sea	[65]
Cambrian (541–485 Ma)	USA	Deformation cracks, intraclast angularity, polygonal plan-view geometries, and intense scouring	Seismic activity	Shallow subtidal intracratonic shelf	[66]
Ordovician (485.4–443.8 Ma)	Poland	Single siliciclastic turbidite lobe,	Unknown	Outer ramp carbonates	[67]
Ordovician (485.4–443.8 Ma)	Canada	Widespread carbonate conglomerate, argillaceous marker beds, and lack of quartz sand/silt (not storm-related).	Seismic activity	Shallow, tropical epeiric (carbonate) shelf within an intracratonic basin	[68]
Ordovician (485.4–443.8 Ma)	Argentina	Graded beds overlying sharply scoured surfaces, anomalous intraclastic rudstone and laminated dolostone, lack of evidence for storms or eustatic sea-level changes.	Seismic activity	Shallow subtidal	[69]
Silurian (443.8–419.62 Ma)	Ukraine	Sand to boulders size clasts and grainstone to mudstone laminae.	Unknown	Tidal flat/shallow-marine carbonates	[70]
Devonian (419.62–358.86 Ma)	USA	Carbonate megabreccia with shocked quartz, ejecta spherules, and multiple graded tsunami units.	Meteorite impact	Outer carbonate platform to slope	[71]
Devonian (419.62–358.86 Ma)	Poland	Stromatoporoid accumulations (allobiostrome and parabiostrome) with distinctive morphometric and taphonomic features indicating deposition by a high-energy extreme event.	Unknown	Shallow-water carbonate platform	[72]
Devonian (419.62–358.86 Ma)	China	Isochronous event deposits (rudstone, calcirudite, turbidite, homogenite) across platform-margin slope and inter-platform trough facies	Bolide impact	Platform-margin slope to inter-platform	[73]
Permian (298.90–252.17 Ma)	Brazil	Extensive soft-sediment deformation (seismites), thick debritic conglomeratic breccia bed with irregular, scoured base and fining-upward structure, chaotically oriented, imbricated clasts (10–400 cm) of chert, siltstone, and sandstone derived from underlying strata, shock-metamorphosed zircon grains.	Bolide impact	Shallow epicontinental marine to lacustrine conditions	[74]
Permian (298.90–252.17 Ma)	Brazil	Shell-rich rudstone and conglomerate beds, bioclastic concentrations with taphonomic signatures of sudden reworking.	Seismic activity	Lacustrine system	[45]
P-Tr boundary (~252 Ma)	India	Fining-upward bioclastic grainstone beds interbedded with argillites, hummocky cross-stratification, grading, and coarse clasts.	Seismic activity	Outer shelf to upper slope marine setting	[75]
Triassic (251.9–201.3 Ma)	Germany	Laterally extensive crinoid-bearing bioclastic bed, Heterogeneous facies association with grain-size sorting and lateral transitions, Chaotic orientation of terebratulids and crinoids, mixed faunas from various ramp zones.	Seismic activity	Outer ramp to basin setting	[76]
Jurassic–Cretaceous boundary (~143 Ma)	France	Basal erosional surface, soft-sediment deformation structures, lateral facies variation, and erosional conglomerate overlain by wood fragments and clay.	Unknown	Coastal to nearshore	[56]
Cretaceous (143.1–66.0 Ma)	Pacific margin	Extraordinary amber concentrations, flame-like structures, terrestrial material (resin) over long distances from continental sources, and Absence of subaerial exposure before deposition.	Massive submarine landslide	Pelagic deep-sea	[77]
K–Pg boundary (~66 Ma)	Argentina	Presence of erosional and reworked horizons and mixed continental and marine fossils.	Chicxulub impact	Shallow marine to marginal marine	[78]
K–Pg boundary (~66 Ma)	Cuba	Ripple cross-laminations, changes in detrital provenance, and maximum grain-size variations.	Chicxulub impact	Shallow marine to coastal shelf	[47]
K–Pg boundary (~66 Ma)	Chile	Single iridium anomaly, Palynofacies disturbance: abrupt transition from fresh cuticles to degraded plant matter and spike in spore abundance.	Chicxulub impact	Fluvio-deltaic setting	[51]
Eocene (56–33.9 Ma)	USA	Sandy carbonaceous clay with carbon glass and rock fragments, sandy-matrix breccia with terrestrial (paleosol rip-ups, wood) and marine clasts (fossiliferous chert)	Chesapeake Bay impact	Shallow marine/coastal plain setting	[79]
Eocene (56–33.9 Ma)	Spain	Typical deposits represent large, disintegrative submarine landslides, but with reduced occurrence during the thermal maximum	Initial Eocene Thermal Maximum (IETM)	Deep-sea continental slope and abyssal settings	[80]
Late Oligocene–Early Miocene (27.82–20.44 Ma)	Southwest Pacific	Massive spilite-rich rudite, graded coarse and fine rudite–arenite, intraformational rudite, brown/grey siltstone, minor calcarenite, airfall tuff, and Bouma-type turbidites, Dominance of massive and graded rudite–arenite beds from high-density turbidity currents.	Volcanic activity and slope failure	Deep-sea fan system	[81]
Miocene (23.03–5.333 Ma)	Chile	Boulder-bearing breccia and poorly sorted sandstone, inverse to normal grading, mixed subaerial, beach, and marine sources.	Seafloor fault	Shoreface	[82]
Miocene (23.03–5.333 Ma)	Hungary	Cobble- to boulder-grade gravel, subangular clasts derived from local Cretaceous sandstone, landward imbrication, mixed clayey-sandy matrix with articulated molluscs and ostracods	Seismic activity	Lacustrine system	[43]
Miocene (23.03–5.333 Ma)	Spain	Folded layers overlain by convex-upward, stratified megahummocks, thick shell-debris bed, erosional surfaces, inflow/backflow structures, and bioclastic redeposition	Seismic activity	Outer-ramp carbonate setting	[11]
Late Miocene–Early Pliocene (11.63–3.60 Ma)	Indonesia	Normal-graded sandstone with disturbed structures (siltstone rip-ups, clay clasts, flame structures), bimodal to multimodal grain-size distribution, mixed marine microfossils (inner to middle neritic).	Seismic activity	Shallow-marine	[83]
Pliocene (5.333–2.588 Ma)	Italy	Composite shell bed, chaotic taxonomic mixing, draping geometry (thicker in lows), articulated valves, limited breakage, reworked deep-infaunal bivalves, and rapid burial	Unknown	Shallow-marine carbonate	[15]
Pleistocene (2.588–0.0117 Ma)	Maio Island (Cabo Verde)	Multiple stacked tsunami conglomerates and sandstones, erosive bases, rip-up clasts, high runups (>60 m a.s.l.), coarse clasts reworked from coastal settings.	Volcanic flank collapses	Coastal setting	[20]
Pleistocene (2.588–0.0117 Ma)	Santiago Island (Cabo Verde)	Erosive base, rip-up clasts of paleosol, high runup (>250 m a.s.l.), marine bioclasts out of life position, poorly to very poorly sorted conglomerate with boulders.	Volcanic flank collapse	Coastal setting	[16]
Pleistocene (2.588–0.0117 Ma)	Algeria	Irregular erosive base, lateral facies variation, wide grain-size range (clay to boulders), normal and inverse grading, angular boulders, fragile and hard-rock rip-up clasts, injection and deformation structures, imbricated and chaotic shell orientations, well-preserved mixed fauna from supralittoral to circalittoral zones, good sorting and sharp-edged fragments.	Seismic activity	Coastal setting	[25]
Pleistocene (2.588–0.0117 Ma)	Spain	Erosive unconformity, random accumulations of Glycymeris and Acanthocardia shells; well-preserved, disarticulated bivalves; and chaotic fabric.	Seismic activity	Coastal setting	[37]
Holocene (11.7–0.0 ka)	Ionian Sea	Thick (up to 10–20 m) graded mud turbidites and megaturbidites.	Collapse of the Santorini caldera.	Deep-sea to abyssal plain	[84]
Holocene (11.7–0.0 ka)	Spain	Erosion of beaches and shoreline retreat, breaching and overwash of sandy barriers, deposition of bioclastic layers, and geomorphological reshaping of estuarine mouths.	Seismic activity	Estuarine and coastal setting	[85]
Holocene (11.7–0.0 ka)	Japan	Multiple sandy and gravelly sub-layers with scoured bases, inverse and normal grading, hummocky cross-stratification (HCS), alternations of sand and mud drapes.	Seismic activity	Shallow marine bay/nearshore	[4]
Holocene (11.7–0.0 ka)	Japan	Marine geochemical and paleontological signatures, erosive basal contacts, fining upward trends, presence of marine fossils, and geophysical (GPR) evidence of lateral continuity.	Seismic activity	Coastal setting	[86]
Holocene (11.7–0.0 ka)	Portugal	Coarse-grained layers with erosive bases, abrupt contrast with background fine sediments, distinct internal structures; element composition anomalies.	1755 CE Lisbon	Offshore shelf	[41]

).

3.1. Precambrian

The oldest tsunami deposits ever reported on Earth dates to Early Archaean (ca. 3.48 Ga) and is located in the Pilbara Craton, Western Australia [59]. The authors link this catastrophic event to local faulting associated with volcanic activity, although an asteroid impact or volcanic eruption is also possible. Field features, include large imbricated clasts, hummocky bedding, and turbidity-type graded layers, indicate bidirectional currents. The lower sedimentary deposits were produced by the incoming tsunami waves, while the upper deposits testify the backwash process. However, previously the Hamersley Group of Western Australia provided the geological record of what was considered until recently the oldest known tsunami (~2.6 Ga), attributed to the Late Archaean–Palaeoproterozoic [61]. Here, sedimentary ripples are preserved in deep-marine shelf deposits containing impact-melt spherules—small glassy droplets produced by asteroid impacts during the Late Archaean. Numerical simulations supporting this interpretation indicate that a bolide with a diameter of 4.9 km, a density of 2.6 g/cm³, and an impact velocity of 20 km/s could generate a crater approximately 60.4 km in diameter. Modelling of tsunami waves produced by the subsequent crater collapse, using shallow-water wave theory, suggests that impacts occurring between 2000 and 8000 km away would have been capable of generating waves strong enough to form the observed ripple structures in the Hamersley Basin. Byerly et al. [60] documented another Archaean example from the Barberton Greenstone Belt in South Africa. Both events are associated with extraterrestrial impacts, which occurred frequently during this period of early Earth history [87]. These impact-related tsunamites contain diagnostic features such as impact spherules, minerals showing shock-induced deformation (e.g., shocked quartz), and distinctive cosmochemical signatures (microtektites) [60,88,89].

During the Mesoproterozoic, within the Changcheng Group (1.8 to 1.6 Ga) of the North China Craton, Wang et al. [63] identified tsunami deposits representing one of the oldest recognized examples of tsunamites in the Northern Hemisphere. This event was attributed to a large earthquake in an active extensional setting during the breakup of the Columbia (Nuna) supercontinent. Although these deposits are approximately 1.5 billion years old, they display sedimentological characteristics essentially identical to those documented from the Quaternary and even from historical to recent tsunami deposits in onshore settings. Key features that distinguish them from normal shallow-marine siliciclastic facies include abundant rip-up clasts, poor sorting with a wide range of grain sizes from clay to gravel, fining-upward sequences indicative of waning flow, reworked and redeposited material eroded from underlying strata suggesting a complex provenance from multiple source units, and erosional scours at the base of beds produced by strong incoming waves and backwash flows. Similar features have been reported from Pleistocene and Holocene outcrops, e.g., [17,20,25] (Figure 3A,B), collectively providing compelling evidence for a high-energy tsunami origin, demonstrating that the fundamental depositional signatures of tsunami events have remained consistent over geological time.

In Western North America, tsunami deposits from a previously offshore marine setting were recognized in the Mesoproterozoic Helena Formation of the Belt Supergroup (~1.45 Ga) [64]. The deposits associated with this event are characterized by the widespread occurrence of molar-tooth structures, which are composed of offshore swept ooids and rounded, coarse-grained, feldspath and quartz sand. These beds display an erosive base that testify to the action of strong, high-energy currents, producing gutter casts and other scour features. Such sedimentary structures indicate deposition under the influence of powerful tsunami-generated flows, attributed to major submarine fault surface ruptures during episodes of tectonic subsidence.

3.2. Palaeozoic

The Palaeozoic (ca. 541–252 Ma) is known as a period of major tectonic and climatic changes, as well as the proliferation of more complex biological forms. Intense tectonic activity created more outcrops and exposures that remain accessible today, which likely contributed to the greater number of published studies on tsunami deposits (~40 works), compared to those documented from the Precambrian (Figure 2). Despite the difficulty in identifying and dating depositional signatures of catastrophic events, evidence suggests that such records (tsunamites and seismites) extend deep into geological time (e.g., Early Paleozoic). For example, Guo et al. [90] documented deformation structures induced by earthquakes of magnitude 7.0–7.6 in the Early Cambrian of western Zhejiang, China, which are classified as seismites. Such strong tectonic activity highlights the potential for tsunami generation during the Early Paleozoic. During the Cambrian, many shallow-marine deposits from both South and North America were traditionally attributed to storm activity [65,66]. However, new interpretation suggests that some of these record tsunami events instead. In the Middle Cambrian La Laja Formation (Argentine Precordillera), coarse-grained strata on an epeiric shelf include lenticular beds of intraclastic conglomerate interbedded within burrowed mudstones. These beds lack stratigraphic patterns, allochthonous input, or evidence of sea-level fluctuation, and instead record brief, anomalous episodes of deep scour and strong oscillatory flow that are best explained by tsunami waves rather than storms [65]. Similarly, the Upper Cambrian Deadwood Formation (Montana, USA) provides one of the earliest confident identifications of tsunamites in the Paleozoic. Here, conglomerates composed of angular to subangular sandstone intraclasts with a high degree of scouring, were long interpreted as storm deposits [91]. However, Pratt [66] proposed recognition criteria that distinguish these deposits from storm-generated counterparts, including deformation cracks, intraclast angularity, polygonal plan-view geometries, and intense scouring, thereby linking them to tsunami activity. These case studies demonstrate that tsunamites can be reliably differentiated from tempestites in ancient shallow-marine successions. Taken together, they challenge the conventional tempestite paradigm and emphasize the importance of recognizing tsunamites as a distinct class of high-energy deposits in the geological record.

Landslides, slumps, and density-based avalanches can generate sediment gravity flows (turbidity current) [92]. According to Shanmugam [29] earthquakes, tsunamis, and cyclones are additional powerful triggers of turbidity currents. A comparable case was reported by Põldsaar et al. [67] from the Middle Ordovician shallow-marine deposits of the Baltoscandian palaeobasin (Northern Europe), where sedimentological evidence indicates a single turbidite event displaying a Bouma sequence. This deposit was interpreted as the product of a rare tsunami, which eroded and transported sediments from land into deeper parts of the basin. Such evidence demonstrates that tsunami-triggered turbidites can also occur in epicontinental settings, not only along active continental margins. Ruban [93] synthesized Ordovician tsunami deposits (485.4–443.8 Ma) from various paleocontinent domains, including Gondwana, Laurentia, Baltica, North China, and Siberia, and concluded that our knowledge of ancient (pre-Quaternary) tsunamis is extremely limited, likely representing less than 0.00001% of all events. Pratt and Sproat [68] documented a widespread carbonate conglomerate in the latest Ordovician Stonewall Formation (Williston Basin, central Laurentia), composed of dolomudstone intraclasts, and interpreted as the result of successive tsunami waves and backwash. This interpretation may also account for other anomalous marker beds in shallow-marine epeiric seas, particularly in settings where storm activity or sea-level fluctuations are unlikely [69].

In Sweden, Parnell [94] concluded that the unusually high concentration of meteorites over a short time interval indicates a flux roughly 100 times greater than present-day rates. These events may have triggered a tsunami during the Middle Ordovician. Concurrently, sedimentary megabreccias (i.e., large rock fragments dispersed in a fine matrix) were deposited in multiple locations. These deposits likely reflect downslope movement of sediments along continental margins on a global scale. The megabreccias may have been triggered by seismic activity and slope destabilization caused by the meteorite bombardment, suggesting that the anomalous occurrence of such deposits in the geological record could serve as a proxy for other episodes of enhanced meteorite delivery to Earth.

Contrary to other systems in the Paleozoic, Silurian tsunami deposits (ca. 443–419 Ma) are still poorly documented. One of the rare examples is reported by Jarochowska and Munnecke [70], who interpreted the Makarivka Member of the Ustya Formation (Ukraine) as a tsunamite, with the run-up phase represented by conglomerate facies, while the heterolith unit composed of grainstone and mudstone laminae reflects alternating high-density landward currents, stagnant intervals that allow mud and land-derived debris to settle, and backwash flows. This study provides rare insights into the interplay between high-energy redeposition and carbon isotope stratigraphy in the geological record, where a tsunami event mixes sediments from different depths and stratigraphic levels, producing clasts with variable δ¹³C values.

In contrast, there are a remarkably large number of deposits related to tsunami events documented for the Devonian period across different domains [71,72,73,95,96]. This abundance may be explained by the fact that the Devonian coincides with major orogenic events, such as the Acadian orogeny (Laurentia-Avalonia collision) and the ongoing Caledonian-Variscan tectonics, which created seismically active continental margins—ideal settings for large earthquakes capable of generating tsunamis. This intense seismic activity has been linked by Du et al. [73] to the formation of seismite structures, including seismic-cracks, sandstone dykes, syn-depositional faults, microfolds (micro-corrugated lamination), fluidized veins, load casts, flame structures, pillow structures, and brecciation, resulting from the collision between the North China Plate and the Qaidam microplate during the Devonian. Extraterrestrial bolide impacts are another trigger of Devonian tsunami deposits as reported by Warme and Kuehner [71] from southern Nevada, where shallow-water carbonate-platform deposits are represented by a carbonate breccia—an unusual sedimentary unit containing shocked quartz grains, unique ejecta spherules, and an iridium anomaly interpreted as impact-generated tsunami event. Additional evidence suggests that the event occurred over a few hours to days, with the impact producing a seismic shock that delaminated the upper 50–100 m of the platform, generating large carbonate blocks. Taphonomic analysis of skeletal concentrations is often used to identify depositional processes in onshore settings [24,25,26,34,97], or from offshore environments [11,15,19]. From the central Poland, within the Devonian Kowala Formation, Łuczyński [72] applied the same methods to examine the sedimentary history of two types of stromatoporoid accumulations—an allobiostrome and a parabiostrome. By combining sedimentological observations with taphonomic analyses, he concluded that the large-scale onshore redeposition of stromatoporoid skeletons could only have occurred through erosion at considerable depths, most likely caused by a tsunami event.

Despite the long duration of the Carboniferous (~60 Ma) and its complex tectonic setting, this system represents a paradoxical interval in tsunami research [93], with only rare scientific papers reporting evidence of deposits attributed to tsunami events, most of them discussed only marginally, e.g., [98,99,100,101,102]. This rarity likely reflects a research bias rather than a genuine absence of tsunamites or their poor preservation. Similarly, what may be the only high-energy extreme event recorded in deposits from the Late Permian from the intracratonic Paraná Basin (Brazil), is documented by Tohver et al. [74] which were interpreted as being related to an impactogenic earthquake.

3.3. Mesozoic

The Permian–Triassic boundary (P–Tr), represents one of the most challenging limits in the geological records, characterized by the largest biological crisis in Earth’s history [103,104], during which approximately 90% of marine species and about 70% of terrestrial species were eliminated [105]. Based on limited occurrences of shocked quartz, Retallack et al. [106] proposed that an extraterrestrial impact was the cause of this event. However, according to Reichow et al. [107] and Korte et al. [108], the eruption of the Siberian Traps flood basalts is currently regarded as the most probable trigger for the extinction event, as the environmental disturbances it generated are widely considered decisive. Within this debate and close to the paleontologically defined P–Tr boundary at Guryul Ravine in Kashmir (India), Brookfield et al. [75] reported sedimentary structures interpreted as seismites, including convoluted bedding and fluid-escape structures, followed by three lenticular, fining-upward bioclastic grainstone beds interbedded with argillites and interpreted as tsunamites. According to the authors, the presence of hummocky cross-stratification and grading within these beds, combined with the associated physical processes and faunal assemblages, indicates deposition by waning irregular waves at water depths exceeding 100 m. The recurrence of such deposits suggests multiple successive tsunami events in a large open-ocean setting. However, the absence of geochemical evidence for an extraterrestrial impact, supports a terrestrial origin, likely related to tectonic or volcanic activity. The same types of sedimentary structures have also been observed in offshore tsunami deposits associated with large earthquakes from Spain [19].

At the P–Tr boundary (~251 Ma), seismites and tsunamites have been reported in multiple sections globally [109,110,111,112,113,114], suggesting that voluminous volcanic activity, earthquakes, bolide impacts, and related tsunamis may have been significant contributors to the environmental stress during the largest mass extinction in Earth history. A recent review of the extant literature on Triassic tsunamis [114], highlighted that most reported examples cluster around two major intervals: the Permian–Triassic transition and the end-Triassic crisis, close to the Triassic–Jurassic boundary. The certainty of these interpretations remains limited, as many of the proposed tsunami deposits could alternatively be attributed to storm activity. This ambiguity arises from two main factors. First, distinguishing between storm- and tsunami-generated facies in the rock record is challenging due to the inherent variability of such sedimentary features. Second, the paucity of detailed sedimentological studies available for the Triassic further complicates matters. Differentiation between tsunami and storm deposits thus relies primarily on the development and application of robust diagnostic criteria rather than the number of examples available from a specific period. Therefore, although the potential role of tsunamis during these major biotic crises is intriguing, the existing evidence should be considered provisional until further investigations, involving more precise sedimentological, stratigraphic, and geochemical analyses, like those reported near the K–Pg boundary, are conducted. The difficulty in unequivocally identifying the triggering mechanism also means that most interpretations of tsunamis-related deposits in the Mesozoic remain uncertain. A representative example is provided by Schnyder et al. [56], who attributed sedimentary units deposited around the Jurassic–Cretaceous boundary from northern France to a tsunami event. Their interpretation was based on comparison with modern tsunami deposits and the recognition of diagnostic features, including a basal erosional surface, soft-sediment deformation, the presence of wood fragments, and clay layers containing a mixed assemblage of continental and marine fauna. This last criterion, faunal mixing across different marine environments or with continental fauna, has often been considered a benchmark indicator of tsunami deposits [15,19,24,25,26,34,37]. Nevertheless, despite the presence of these diagnostic features, the absence of direct evidence for the triggering source led the authors to classify these sediments as possible rather than definitively tsunamites.

Since the Cretaceous, and in contrast to the other systems of the Mesozoic, research on palaeotsunami deposits has expanded considerably. More robust interpretations are now supported by multiple lines of evidence, including several diagnostic criteria. These include anomalous concentrations of amber in deep-sea settings far from its continental source and the occurrence of deformation features identical to soft-sediment structures in surrounding deposits [77], rapid burial of mixed faunas without time for reworking, taphonomic contrasts such as articulated shells versus fragmented shells, the intermixing of freshwater and marine faunas within the same layer [115], and the exceptional preservation of molluscan fossils from different shallow-marine settings [116]. Additional diagnostic features include the large lateral extent and facies uniformity of the deposits, their multi-scale facies heterogeneity, and their predictable petrophysical properties [117]. The latter case-study highlights how tsunami deposits can be identified not only through sedimentological and taphonomic features, but also by their large-scale architecture and internal facies organization, with implications that extend beyond sedimentology (e.g., for reservoir characterization).

The triggering mechanisms of these deposits are linked to a range of geological processes, including major earthquakes [118,119], submarine landslides [77], and even bolide impacts [120]. This diversification of recognized triggers reflects not only the improved preservation potential and accessibility of Cretaceous tsunamites, but also the increasing application of modern sedimentological and palaeontological approaches, which enable a more confident distinction between tsunami deposits and other high-energy event beds, such as storm deposits.

3.4. Near the K–Pg Boundary

The search for the ultimate cause behind the extinction of the dinosaurs and nearly 72% of all other species stands as one of the most widely debated questions in Earth’s history [121]. About sixty-six million years ago, a bolide impacted at Chicxulub on the northwestern corner of the Yucatán Peninsula, Mexico [78,122,123,124,125]. This impact is now largely accepted as the primary cause of the mass extinction that marks the Cretaceous–Paleogene (K–Pg) boundary ([125] and references therein). It generated a complex cascade of hydrodynamic and mass-transport processes around the Gulf of Mexico and beyond, including tsunamis, impact-seiche waves, margin collapse, and sediment-gravity flows [50].

The K–Pg boundary provides a particularly compelling case for studying tsunamis generated by catastrophic extraterrestrial impacts. Consequently, tsunami deposits from this interval have been the focus of extensive research worldwide and studied for more than 30 years [97]. Well-documented examples are found across the Gulf of Mexico and Caribbean regions, including the Brazos River in Texas, northeastern Mexico, Cuba, and Haiti [49,126,127,128,129].

In Western Cuba, the Moncada Formation preserves a tsunamite characterized by shocked quartz, impact glass, an iridium anomaly, and complex ripple cross-lamination interpreted as evidence of oscillatory tsunami palaeocurrents [47]. Similarly, in the La Popa Basin of Northeastern Mexico, an 8-m-thick chaotic unit contains Chicxulub ejecta overlain by graded sandstones that reflect multiple tsunami surges and hyperconcentrated flows generated by slope failure [49]. High-altitude and continental deposits at Tanis (ND, USA) have provided complementary evidence, when a remarkable seiche-induced inland surge deposit containing fish fossils with impact spherules embedded in their gills, capturing the near-instantaneous arrival of seismic waves and ejecta fallout [50]. In Patagonia, an iridium anomaly, platinum group enrichment, and palynofacies changes at the boundary horizon further attest to the impact signal [51].

Identifying tsunami-related deposits in offshore settings is challenging due to limited access to shelf archives [22]. Preservation potential strongly depends on the depositional environment, sandy tsunami deposits in areas shallower than ~65 m are particularly prone to storm reworking [39], providing a framework to assess which K-Pg deposits are likely to be preserved. However, another deep-water record from the southeastern Gulf of Mexico (DSDP Sites 536 and 540) reveals a five-unit succession at the boundary, interpreted as deposits of a giant wave and associated backflow in a basin-floor to lower-slope setting [130]. These cores are characterized by graded sand, rip-up clasts, and reworked bioclasts, which extend the tsunami signal into deeper parts of the basin and demonstrate rapid, high-energy emplacement immediately following ejecta deposition. In Argentina, Scasso et al. [48] documented a coarse-grained sandstone bed within the Jagüel Formation showing erosive bases, grading, and hummocky cross-stratification, features interpreted as products of tsunami flows in a distal neritic setting.

Following the impact-extinction hypothesis, tsunami deposits became central to testing and documenting catastrophic sedimentary responses to the Chicxulub event, and subsequent resistance to the impact theory stimulated detailed sedimentological, geochemical, and paleontological investigations [121]. Hydrodynamic models further confirm the extraordinary reach of the Chicxulub tsunami, with velocities capable of eroding sediments at distances exceeding 10,000 km from the crater [131]. Collectively, these records demonstrate that K–Pg tsunamites are globally distributed and diverse in sedimentological character, ranging from chaotic breccias and graded marine sandstones to inland surge deposits. Their identification relies on the integration of sedimentology, paleontology, geochemistry, and geophysical modelling, and these collectively provide compelling evidence that the Chicxulub impact tsunami was not only a local catastrophe but also a globally significant sedimentary phenomenon that played a major role in the End-Cretaceous mass extinction.

3.5. Paleogene and Neogene

Compared with K–Pg, tsunami deposits from the Paleogene are fewer and patchier, reflecting both preservation and recognition challenges as well as research bias toward younger successions. Despite this, several major tectonic phases occurred worldwide during this System [132,133,134,135,136], many of which strongly influenced sedimentation, basin evolution, and even potential tsunami generation. Documented Paleogene tsunamites are concentrated in the Eocene, when two large impacts (Chesapeake Bay, USA) occurred within ~0.5 Myr of each other [79,137,138]. In the Atlantic Coastal Plain of North Carolina (USA), a sandy-matrix breccia with mixed terrestrial (paleosol rip-up clasts, petrified wood) and marine (fossiliferous chert, meter-scale clasts) components, interpreted as high-energy tsunami surges and a capping quartz sand with gravel, interpreted as a subsequent tsunami wave or backwash [79]. This extreme event related to Late Eocene Chesapeake Bay impact illustrates the preservation of coupled impact ejecta and tsunami deposits in a shallow-marine to marginal-terrestrial setting. From the New Jersey continental margin at ODP Sites 903 and 904, the same impact generated both strong ground motion and tsunami waves [137,138]. These impact-related tsunamites highlight the importance of extraterrestrial forcing, whereas other Paleogene catastrophic events underscore tectonic and climatic trigger.

Climatic perturbation also played an indirect role in triggering tsunami. Rapid warming events, such as the Initial Eocene Thermal Maximum (IETM), which represents the warmest period on Earth during the Cenozoic [139], influenced sedimentary processes by affecting slope stability and promoting landslide and turbidity current [140,141]. When slope failure occurs, the sudden pressure drop or sediment remobilization can destabilize hydrates within or beneath the failed mass, leading to rapid dissociation and the release of methane [142,143]. This process not only contributes to weakening slope sediments, thereby enhancing the likelihood and magnitude of failure-generated destructive tsunamis that can produce a turbidity current activity on non-glacially influenced margins [80,144,145]. A similar case is documented from the Eocene of Zumaia in Northeastern Spain, demonstrating that such findings are highly relevant for assessing future landslide-triggered tsunami hazards and the vulnerability of seafloor infrastructures, highlighting the value of recent geological records for recurrence estimates [80,145,146,147,148].

An earthquake may generate tsunami waves that subsequently induce turbidity currents, may trigger turbidity currents directly without the involvement of a tsunami [29], or in the case of large submarine landslide, turbidity current and slope failures may themselves act as the primary mechanism for tsunami generation. This overlap often leads to confusion between triggering mechanisms and depositional processes, as illustrated by the Late Oligocene–Early Miocene from the Lower Sarava Formation in the Southwestern Pacific. Deposited within a deep-sea fan at depths > 4.25 km and fed by volcanic material from the Vitiaz arc, its thick rudite–arenite beds have been interpreted as tsunami-related deposits [81]. This interpretation is supported by their unusual thickness, coarse grading, and rapid emplacement, features consistent with high-energy flows most likely triggered by strong seismicity, volcanic activity, or slope failure within the interarc basin.

The research on Neogene (23–2.58 Ma) tsunami deposits has progressed considerably. These deposits have been identified in a wide range of settings, from onshore to offshore environments [11,15,19,82,83,149,150,151,152,153,154,155,156,157], as well as within lacustrine systems [43]. Multi-proxy approaches are commonly used, with sedimentary evidence documenting both the run-up and the backwash phases.

From the Miocene Makran accretionary complex (Iran), a chaotic unit is represented by reworked blocks of various lithologies and sizes dispersed within a muddy matrix, previously interpreted as a tectonic mélange diapirically emplaced from depth [158,159]. According to Burg et al. [160], the presence of ophiolitic blocks and reworked fragments of underlying turbidites, together with soft-sediment deformation structures within this unit, supports a sedimentary rather than tectonic origin, making it comparable to large debris flows observed along modern continental margins or unstable volcanic edifices. Another debated case is reported from northern Chile, where Miocene shoreface sandstones of the Caleta Herradura half-graben were firstly interpreted as deposits formed by tsunami surges [82]. The succession consists of two erosively based units, likely emplaced by successive waves, and displaying diagnostic features attributed to tsunami backwash processes. These include exceptionally coarse grain size compared to surrounding deposits, sharp erosional bases, mixed sediment sources, abrupt lateral facies changes, and grading from chaotic to massive textures. However, this interpretation was later re-evaluated by Bahlburg et al. [161], who argued that the deposits are better explained as debris flow units representing physically continuous distal subaqueous equivalents of graben flank alluvial fan breccias, rather than tsunami deposits. For Son et al. [157], discriminating specific triggering events from sedimentary gravity-flow deposits is particularly challenging, especially in ancient records where diagnostic features are often subtle, overprinted, or poorly preserved. Such sedimentary features may be produced by both tsunamis and the other type of events mentioned above; therefore, distinguishing between the triggering mechanism and the depositional processes is crucial to avoid misinterpretation [29].

Nevertheless, despite such interpretative challenges, particularly in offshore settings, the Miocene deposits of Spain [11] and the Pliocene successions of Italy [15] provide well-preserved examples that, together with model cases, serve as methodological benchmarks. These studies highlight the value of multi-proxy approaches in improving the reliability of interpretations and in distinguishing tsunami deposits from other high-energy events such as storms and mass-flow processes. The Hackcheon Chogok Megaturbidite from the Miocene Pohang-Youngouk basin (Republic of Korea) represents one of the best-documented examples where sedimentological, structural, and provenance evidence point to a tsunami-triggered megaturbidite [157], helping to distinguish it from other, similar looking gravity flow deposits in the geological record. Recently, Di Celma et al. [156] reported that in the Lower Miocene Chilcatay Formation of southern Peru, offshore tsunami deposits are characterized by several key features: an irregular basal erosion surface with multiple asymmetric and symmetric scours, unusually coarse grain-size compared to the surrounding background sediments, absence of internal bioturbation, the presence of multiple graded beds, and a mixture of both well-rounded and angular clasts derived from nearshore and subaerial coastal areas. Additionally, the abundance of vertebrate skeletal remains (sedimentological concentration sensu Kidwell et al. [162]) embedded within these deposits suggests that such catastrophic tsunami events could be responsible for mass mortality of fishes, turtles, and dolphins, as observed in modern tsunami events [82]. The preservation potential of tsunamites in offshore settings is enhanced by their emplacement below the storm wave base, where they remain largely unaffected by typical wave reworking. For Puga-Bernabéu and Aguirre [19], shell beds generated by tsunami backwash are characterized by a distinctive taphonomic signature, including the admixture of organisms from different ramp zones, reduced fragmentation, sharper shell edges, oblique to perpendicular skeletal orientations, chaotic stacking of concave shells, and a lower incidence of encrustations and borings.

Within the Late Miocene–Early Pliocene Bentang Formation at Tegal Buleud (Indonesia), where active subduction between the Australian and Eurasian plates frequently generated strong earthquakes, the first and oldest geological evidence of a tsunami has recently been identified in the region [83]. This discovery not only extends the tsunami record of Java and Indonesia into the Neogene but also confirms some benchmarks usually considered the most critical criteria marking tsunamites across the Phanerozoic, such as clay clasts, siltstone rip-up clasts, normally graded sandstones, bimodal to multimodal grain-size distributions, and mixed marine fossils from different environmental settings (e.g., inner shelf to middle neritic). Another criterion, reported by Le Roux et al. [163] within deposits linked to a large tsunami event affecting the southern Chilean coastline during the Pliocene, includes inverse grading, planar laminations, ripple or trough cross-lamination, and the injection of sand-rich material into the underlying cohesive substrate. Similar characteristics have been documented in tsunami-related deposits from foreshore and backshore settings in the Mediterranean [25] and the Atlantic Ocean [16,17,20] (Figure 4).

3.6. Pleistocene

In contrast to other Phanerozoic times, the Pleistocene (2.588–0.0117 Ma) represents the most crucial interval for understanding palaeotsunami deposits. This epoch provides the most continuous, accessible, and well-preserved sequences, which can be directly compared with historical and recently documented events across diverse environmental settings. Interest in palaeotsunami deposits from this interval has grown not only because of their significance for reconstructing palaeocoastline evolution, palaeogeomorphology, and reservoir characterization [26,117], but also because studies of Quaternary tsunamis highlight the history of volcanism and tectonic activity, particularly earthquakes and plate-boundary processes, essential for hazard assessment. Many Pleistocene tsunami-related deposits (tsunamites), originally triggered by earthquakes or landslides, were later overlain by additional tsunami deposits of the same age [17,20,164,165] or by younger Holocene and even historical events [25,37,38]. This stratigraphic superposition indicates that seismic activity and tsunami generation persisted from the Pleistocene into the Holocene and historical periods. However, as several previously cited examples demonstrate, tsunami events in this interval were not exclusively linked to tectonic activity; flank collapses and other non-seismic triggers in tectonically less active regions also played a significant role. Together, these records underscore the long-term recurrence of tsunamigenic processes across a range of geodynamic settings.

The Pleistocene Epoch was characterized by repeated glacio-eustatic sea-level fluctuations, tectonic activity that generated marine terraces, and large-scale slope instabilities [25,166,167,168,169,170,171,172,173,174]. These conditions were highly favorable for tsunami generation. Several well-documented Pleistocene examples highlight the diversity of depositional settings associated with catastrophic events, often linked to seismic activity and volcanic collapses. Such occurrences have been reported worldwide, including the Pacific Basin—particularly Japan [175] and New Zealand [176,177], as well as in the Atlantic regions [16,17,20,178] and the Mediterranean [25,34,37,84].

Oceanic volcanic islands are prime settings for investigating tsunami deposits, as volcanism, flank instability, erosion, and vertical crustal movements (subsidence or uplift) have driven rapid morphological changes during the Quaternary [17,168,179,180,181]. Compared to regions located along active subduction zones, such as the margins of the Pacific and Indian Oceans, tsunami hazards in the NE Atlantic archipelagos, including Madeira, the Canary Islands, and Cabo Verde, may be significantly underestimated [20]. While subduction zones are recognized as the primary sources of large, destructive tsunamis, oceanic volcanic islands can also generate catastrophic events through mechanisms such as massive flank collapses, volcanic explosions, and associated landslides. These mechanisms can generate a mega-tsunami [17]. The term mega-tsunami refers to a wave amplitude exceeding 50 m [182]. In comparison, the 1958 Lituya Bay tsunami [183] is often cited as the only historical mega-tsunami, with a maximum run-up of 524 m, however, its effects were spatially limited and rapidly diminished away from the landslide source [17]. The limited number of historical tsunami records in the Central Atlantic may create a false sense of security; however, geological evidence demonstrates that these islands have experienced multiple large-scale events during the Pleistocene [16,24,164,165,184].

One of the best examples interpreted as evidence of a mega-tsunami generated by catastrophic volcanic flank collapse is represented by chaotic marine conglomerates and large boulders found at a wide range of elevations on the flanks of the Cabo Verde islands [16,20,164,165,178] and Canary Islands [184,185]. These deposits, often perched tens to hundreds of meters above present sea level, display sedimentological features consistent with high-energy emplacement. Their interpretation as tsunami deposits was previously debated [17]. Previous studies proposed that these accumulations could instead represent uplifted littoral deposits [186,187,188,189]. However, more recent investigations integrating detailed sedimentological, geomorphological, and chronological data [17,18,19,20] have provided robust evidence supporting their interpretation as tsunami deposits. These advances highlight the importance of combining multiple lines of evidence to accurately distinguish tsunami deposits from uplifted coastal facies in island settings.

Dating tsunami deposits using methods such as U-Th, radiocarbon and/or amino acid racemization in shells, cosmogenic exposure techniques, or stratigraphically constraining the deposits between the underlying and overlying units is critical, as it not only constrains the timing of extreme events but also helps to evaluate whether they had a regional impact. By establishing the timing of events at multiple sites, it is possible to correlate deposits across the region and assess the extent and impact of individual tsunamis. This is particularly relevant for the Last Interglacial (MIS 5e), when tsunami deposits have been reported across the Western Mediterranean. In Southwestern Spain, MIS 5e deposits at La Mata and Torrevieja reflect coastal uplift that occurred between MIS 7 and MIS 5e [37]. These successions include accumulations of randomly arranged and well-preserved Glycymeris and Acanthocardia mollusc bivalve shells, as well as the transport of enormous boulders onto stepped abrasion platforms. These features are interpreted as tsunami surge deposits generated by an offshore earthquake. Comparable successions interpreted as the result of an offshore earthquake-induced tsunami are also documented from Northwestern Algeria [25], where the Glycymeris-rich Unit of the Hachacha Plateau is characterized by mixed molluscan assemblages, including the so-called Senegalese fauna, excellent shell preservation, and diagnostic sedimentological features such as erosive bases, lateral facies variations, wide grain-size ranges (from clay to boulders), rip-up clasts, and soft-sediment deformation structures. Both Spanish and Algerian examples emphasize the importance of precise dating for identifying synchronous tsunami events, and their shared faunal content (notably the dominance of Glycymeris), combined with geomorphological, sedimentological, and taphonomic characteristics, suggest a regional-scale impact during MIS 5e, directly linked to seismic activity along the Africa–Europe plate boundary.

From Central Japan, anomalous deposits are interpreted as earthquake-induced tsunami deposits from the Pleistocene [175]. These show a rare example where successions collectively suggest a three-stage depositional model, with the preservation of a lower unit that shows the presence of contorted structured beds (convolute structures) interpreted as earthquake event deposits (seismites), which may testify to the triggering mechanism of this catastrophic event or help to relate these tsunamites directly to seismic activity. The middle unit suggests tsunami current-dominated processes, whereas the upper unit represents a suspension-settling event characterized by the presence of cross-laminated sand with mud drapes. The same succession has been reported within tsunami deposits related to seismic activity documented from northwestern Algeria [25], where seismite structures were found near the erosion surface. In this case, the authors described soft-sediment deformation structures (SSDS) rather than convolute bedding, and a mud layer instead of the mud drapes. The features reported above, including the presence of unit presenting seismite structures, overlain by tsunami deposits, which are themselves characterized by rapid lateral facies variation, changes in thickness, and in some cases the presence of large boulders and shell beds, can therefore be used as indicative of an earthquake-induced tsunami sequence in similar tectonically active settings.

4. Holocene to Recent Times Records

The Holocene (the last ~11,700 years) is characterized by relatively stable coastlines, following the last major post-glacial sea-level rise. This stability has created ideal conditions for the preservation and dating of palaeotsunami deposits [190,191,192]. The good preservation and dating of these deposits allow for direct correlation with historically documented events. Their significance extends far beyond academic interest, as they are fundamental for extending the often short and incomplete record of tsunamigenic events in any given region [18,22,33,37]. Moreover, the study of tsunami deposits from the Holocene to recent times enables the scientific community to better characterize the full range of diagnostic features of these deposits by applying the principle of actualism [193]. This approach also strengthens the assessment of geological hazards, particularly those triggered by seismic activity [37].

From the Holocene, the identification and interpretation of tsunami deposits are based on an integrated combination of multi-proxy approaches, including sedimentology, paleontology/micropaleontology (e.g., foraminifera, diatoms, mollusks, corals, and other fauna), geochemistry, and geomorphology with precise dating techniques comprising radiocarbon (¹⁴C), optically stimulated luminescence (OSL), and amino acid racemization (AAR) [12,34,35,37,40,194,195,196,197,198,199].

Several studies have emphasized the value of estuaries as natural archives for tsunami research, since they can preserve a wide range of geological evidence [85,199,200]. The estuaries are excellent coastal systems for recording tsunami deposits because they preserve a variety of geological indicators, including beach erosion and shoreline retreat, breaching and overtopping of sandy spits, and the deposition of bioclastic layers above the normal action of sea waves and storms [85]. The multiple fining layers or units reflect multiple successive waves during the extreme events (tsunamis) [15,25,82]. These features have also been reported from the 2004 Indian Ocean tsunami [201], where the catastrophic event generated multiple patterns in vertical textural trends represented, from bottom to top, by upward fining, upward coarsening, and uniform units. In the same context, from eastern Japan, a Holocene depositional model demonstrates that, while individual structures such as hummocky cross-stratification and graded bedding resemble storm deposits, the vertical stacking forms a distinctive signature of tsunami wave trains [4]. This interpretation allowed the authors to propose a genetic vertical facies model organized into four units (Tna–Tnd), reflecting changes in wave size and energy, and providing a key criterion to distinguish tsunamites from tempestites in the same setting. However, this interpretation has been challenged. For Shanmugam [29], a single depositional event, such as a tsunami cannot reasonably account for the formation of both traction carpets and hummocky cross-stratification (HCS) based on hydrodynamic consideration. This is because these features form via genetically unrelated processes, dispersive pressure and oscillatory flow, respectively. This controversy remains a challenge not only in distinguishing storm from tsunami deposits, but also in determining whether these succession units are related to a single or to multiple tsunami events that occurred close together or separated in time.

Many terms such as turbidite, megaturbidite, debrite, gravitite, and homogenite have been used with overlapping meanings and are often collectively referred to as tsunamites [29]. A classic example is found in the eastern Mediterranean, near the Calabrian and Hellenic arcs. There, a thick marl layer composed of clay, quartz, nannofossils, and foraminifera has been described. This thick unit is characterized by a sharp basal contact, a normally graded sandy base, and overall homogeneity-showing no visible sedimentary structures in core sections or X-radiographs. Since its discovery, it has been interpreted as a homogenite [202]. According to Cita et al. [84], this deposit is a key piece of evidence for the tsunami generated by the Late Bronze age collapse of the Santorini caldera (~3500 years ago). The event triggered large-scale slope failures that, in turn, generated sediment gravity flows. These flows deposited turbidites and megaturbidites in the abyssal plains of the Ionian Sea, while the distinctive homogenite was laid down across the subdued topography of the Mediterranean Ridge, an area often referred to as Cobblestone Topography. This interpretation suggests that tsunamis, turbidites, mass flows, and slope failures are closely interconnected processes. A comparable relationship has been reported from the northern flank of Tenerife, where a massive volcanic collapse mobilized a volume of ~200 km³ offshore, producing three debris lobes and seven turbidites, and was associated with a Pleistocene mega-tsunami [17,203].

The study of both historical and modern tsunami deposits, particularly those from the Lisbon earthquake tsunami (1755), Indian Ocean tsunami (2004), and Tohoku-Oki tsunami (2011), has served as a crucial calibration tool for interpreting older events [18,204,205,206,207,208,209,210,211,212,213]. These modern analogues allowed refining diagnostic features, such as landward fining, inverse-to-normal grading, the presence of intraclasts (rip-up clasts), and the transport of marine sediments (sand, shells, microfossils) far inland [214,215]. Geochemical signatures, including elevated concentrations of salt-derived ions (e.g., Cl⁻, Na⁺, S) in soil and groundwater, have been established as a reliable indicator of seawater inundation, even when the sedimentary evidence is subtle [18,216,217,218,219]. Furthermore, taphonomic analyses of molluscan and foraminiferal assemblages in modern tsunamites have confirmed faunal mixing across different ecological zones as a key indicator [195,220,221,222].

Well-documented Holocene tsunami records have been established in numerous tectonically active regions worldwide. In the Pacific Rim, extensive research has been conducted along the coasts of Japan [35,86,223,224], Chile [225,226,227], the Cascadia subduction zone of North America [228,229,230], and New Zealand [176,231]. In the Atlantic Ocean, the 1755 Lisbon tsunami serves as a benchmark event, with its deposits identified in Spain, Portugal, and Morocco [211,212,218,219,232,233,234,235,236,237,238,239]. The Mediterranean Sea, a basin with a long history of seismic and volcanic activity [240], also preserves a rich Holocene tsunami record, with studies in Turkey [241], Greece [242,243,244], Italy [245,246], the Levant [247,248], Malta [249], Algeria [14,38], and Spain [250] revealing repeated large-scale events.

Despite their higher preservation potential, interpreting Holocene tsunami deposits still presents challenges. These include distinguishing the deposits of closely spaced events [229], differentiating tsunami signals from intense storm deposits (tempestites) in areas where both occur [27,33,251], and understanding the complex effects of local geomorphology and anthropogenic activity on sediment deposition and preservation [252,253]. Nevertheless, the Holocene record, including tsunamis from the last few decades, remains an indispensable reference for interpreting older events. It provides a long-term perspective on tsunami recurrence, magnitude, and source mechanisms, thereby enabling more accurate association between triggers and depositional outcomes, and supporting robust hazard and risk assessments [254,255,256].

5. Tsunamites vs. Tempestites: Key Challenges

5.1. Sedimentological Criteria

Tsunamites and tempestites are sedimentary deposits formed by extreme wave events, such as tsunamis and major storms, respectively. These events are powerful geological agents capable of transporting and depositing vast amounts of sediment both inland and across shallow marine environments [4,17,25,29,57,181,257]. However, distinguishing between tsunami and storm deposits remains one of the most persistent and challenging problems in sedimentary geology [25,28,34]. Both events occur in similar environmental settings (e.g., coastal plains and shallow marine areas) and can share several sedimentary characteristics, such as erosive bases, coarse-grained layers, and internal stratification. Misidentification can therefore lead to significant errors in interpreting the geological record, for example, by attributing a seismic history to a region based on deposits that were actually formed by storms, or vice versa. In fact, several sedimentological and palaeontological characteristics have recently been used to discriminate between these two extreme events across the Phanerozoic, with examples provided from different oceans [12,20,24,25,26,27,37,241].

From the Storegga tsunami deposits in Norway, Dawson et al. [2] were the first to describe the presence of a sharp erosional contact at the base of tsunamigenic layers [198]. Since their pioneering work and following their contribution, this type of erosion has been observed later at the base of deposits related to tsunami events from Japan [258], Papua New Guinea [259], Thailand [260,261,262], and Indonesia [263]. However, Morton et al. [27] reported that this feature can also occur in storm deposits. This overlap highlights one of the major challenges in distinguishing between these two types of extreme events [85], and underscores the need for a multi-proxy approach when interpreting the deposits left by this type of events. The erosional surfaces produced by the 2004 Indian Ocean Tsunami in Tamil Nadu, India, were described as an “erosional unconformity” by Bahlburg and Weiss [264]. This terminology is stratigraphically misleading, as a true unconformity classically represents a significant gap in the depositional record, often involving millions of years [265,266]. Since the 2004 event lacks such a substantial temporal hiatus, its erosional features should not be conflated with stratigraphic unconformities [29]. Instead, such short-lived, high-energy features are more accurately classified as “event surfaces” to prevent confusion in sequence stratigraphic interpretation.

Beyond sharp erosional contacts, other erosive features have also been reported in tsunami-related successions. From northwestern Algeria, within deposits interpreted as the result of a catastrophic tsunami event, Doukani et al. [25] described an irregular erosional surface separating the event deposits from the underlying bedrock (Figure 3F). Here the authors linked this irregular erosion to the lithology and heterogeneity of the eroded basements. According to Bryant and Young [267] and Aalto et al. [268], the bedrock sculpting is related to a tsunami activity. However, Bourgeois [87] (p. 18) has challenged this interpretation and emphasizing that “There is no fundamental basis for the argument that tsunamis are more powerful sculptors than storm waves”.

Tsunami deposits can show a wide grain-size spectrum in sediments, ranging from clay to boulders [20,25,198] (Figure 3C–E). Such conditions are characteristic of catastrophic depositional events, during which extreme hydrodynamic energy rapidly entrains, transports, and reworks sediments from diverse environments in a chaotic manner [269]. According to Bourgeois [87], the grain-size distribution within tsunami deposits is controlled primarily by the availability of sediment sources rather than by the hydraulic competence or flow direction. In many cases, the presence of imbricated boulders is considered another benchmark criteria of tsunami deposits [14,16,17,20,25,270,271,272,273,274], reflecting rapid and high-energy depositional processes. This contrasts with deposits linked to storm events, where sediments are formed under more gradual and oscillatory flow conditions.

Both fragile soil fragments and hard-rock rip-up angular clasts have been reported from the oldest known tsunami event on Earth [59] to more recent records across diverse environmental settings, ranging from coastal lowlands to deep-marine environments [17,20,25,27,79,82,126,130,185,259,275,276,277,278]. These features are typically concentrated in the lower portions of tsunami deposits and along erosional surfaces, particularly where fine-grained sediments are present. In contrast, they are generally absent or rarely documented in storm deposits [27]. Sediment injections, resulting from the extreme dynamic pressure of tsunami surges, provide another key criterion for identifying tsunamites, as they are commonly emplaced into both soft and consolidated underlying strata and reported from several tsunami deposits (Figure 4) [17,20,25].

The identification of tsunami deposits based only on internal sedimentary structures remains a significant challenge. Several tsunamites display features such as cross-stratification, parallel and oblique lamination, hummocky cross-stratification (HCS), and mud drapes [4,25,27,62,145,259,277,279,280,281]. However, many of these structures, particularly parallel and cross-lamination, as well as HCS, are also widely documented in storm-related deposits [281,282], reducing their diagnostic reliability. For instance, parallel lamination has been identified in deposits from the 2004 Indian Ocean tsunami in Sri Lanka and Tamil Nadu [29,201]. Similarly, hummocky cross-stratification has been reported in tsunami deposits from the Precambrian of Australia [61], the Cretaceous–Paleogene boundary in Argentina [48], the Neogene of Spain [13], the Pleistocene of Algeria [25], and the Holocene of Japan [4]. Low-angle cross-stratification has also been described not only in fossil and recent tsunami deposits [15,283] but also in estuarine successions and tidal facies [284]. Unlike tsunami deposits, storm deposits frequently display a wider variety of sedimentary structures beyond simple planar stratification [27]. These include foreset bedding [258,285,286,287,288], backset lamination [289], and climbing ripple structures [282,283,284,285,286,287,288,289,290,291].

In addition to cross-stratification and lamination, both normal and reverse grading have been reported from tsunami deposits at multiple localities worldwide [15,17,20,25,292]. According to Massari et al. [15], the occurrence of these grading patterns reflects sedimentation from prolonged, high-energy flow conditions, which can produce both upward fining and upward coarsening depending on the turbulence and sediment supply. Spiske and Jaffe [293] further emphasized that inverse grading is commonly associated with deposits from tropical cyclones, where oscillatory storm surges and strong backwash can generate upward coarsening trends. By contrast, normal grading has frequently been reported within tsunami deposits attributed to deep environment beneath the storm wave base [83,84], reflecting the waning energy conditions typical of decelerating tsunami flows.

Within this ongoing debate, it is evident that no single sedimentary structure can serve as a definitive indicator of tsunami deposits. Instead, recent studies emphasize that a more reliable identification requires the integration of multiple diagnostic features. These include the association of graded bedding with mud drapes, the presence of rip-up clasts, layers enriched in heavy minerals, sediment injections into both hard and soft substrates, and evidence of both landward and seaward-directed transport, often considered alongside geomorphological context and macro to microfossil assemblages [17,20,24,25,27,37,43,216]. Such multi-proxy approaches substantially improve the reliability of distinguishing tsunamites from tempestites and highlight the importance of integrating sedimentological, palaeontological, geochemical, and geomorphological lines of evidence in paleo-tsunami research (

Table 2. Tsunami versus storm deposit features based on sedimentological, palaeontological and taphonomic, geomorphological, and geochemical criteria.

Criteria	Tsunami Deposits (Tsunamites)	Storm Deposits (Tempestites)	Key References
Trigger and energy source	Generated by long-period seismic sea waves (earthquakes, landslides, volcanic activity), extremely high energy with multiple run-up and backwash phases.	Generated by short-period oscillatory waves and storm surges (cyclones, hurricanes, etc.), energy concentrated nearshore.	[57,265]
Erosional features	Sharp, irregular erosional bases, deep scours and event surfaces, erosion may cut into bedrock or soil, extensive inland erosion.	Shallow scours and erosional truncations near beach/shoreface, usually limited to berms and dunes.	[2,25,27,264]
Grain-size	Very wide spectrum (clay to boulders), often poor sorting, rapid energy fluctuations, possible imbricated boulders.	Moderate to good sorting; coarser sands and gravels typical.	[16,20,25,270]
Sedimentary structures	Combination of HCS, parallel and cross-lamination, normal and reverse grading, mud drapes; stacked sublayers reflecting multiple waves.	Dominated by HCS, cross-bedding, ripple lamination, foresets, climbing ripples; grading mainly normal or inverse depending on wave energy.	[27,190,294]
Rip-up clasts and sediment injections	Common, angular rip-up clasts and injections of sediment into underlying strata from extreme pressure.	Rare, weak injection or absent rip-up clasts.	[17,20,25]
Thickness and lateral extent	Thick, laterally continuous sheets; often thinning landward but extensive over kilometers; multiple subunits possible.	Thinner, lenticular layers restricted to nearshore bars and back-beach environments.	[27,207,261]
Facies variability	Strong lateral variation depending on topography and inundation pathways; may overlie non-marine units.	Typically confined to marine or shoreface settings; limited facies variability.	[20,25,261]
Macrofossil assemblages	Mixed allochthonous assemblages from multiple habitats (marine, continental); high richness and evenness, The abundance of large, heavy shells in molluscan assemblages	Autochthonous or parautochthonous assemblages from local habitats; lower richness.	[12,19,24,25,35,37]
Microfossils (foraminifera, diatoms, etc.)	Mixed assemblages of deep, shallow, and terrestrial taxa; offshore species transported inland.	Restricted to local shallow-marine assemblages; absence of deep-water species.	[295,296,297]
Taphonomy	Angular fragmentation, shell breakage, chaotic orientations (often >30–60°) with dominance of vertical and oblique orientations; evidence of strong backwash; limited bioerosion, low encrustation.	Rounded fragmentation, ordered shell orientation, concave-up stacking, higher bioerosion and encrustation, concave-up, subhorizontal stacking (<30°).	[15,19,25,37]
Geomorphological	Deep scour channels, stripped soil layers, uprooted vegetation, boulder ridges, landward sediment lobes extending far inland (hundreds of meters–km).	Washover fans, beach ridges, dune breaches, limited to low-lying coastal zones.	[28,57,263]
Inland extent of erosion and deposition	Can extend several kilometers inland and tens hundreds of meters above sea level.	Typically confined to nearshore and back-beach areas (<few meters inland).	[259,298,299]
Geochemical signature	Enrichment in Sr, Ca, Mg; marine-derived saline and carbonate influx; distinct XRF peaks; possible marine environmental DNA (eDNA).	Less marine element enrichment; composition similar to local sediments; absence of eDNA marine signals.	[209,217,300]

).

This interpretation is further strengthened by incorporating additional sedimentological and taphonomic criteria, including a wide grain-size spectrum from clay to boulders, pronounced lateral variations in facies and thickness, and the presence of reworked biological remains. Taken together, these features provide compelling support for a high-energy tsunami origin rather than a storm-related process (Figure 5).

5.2. Paleontological and Taphonomic Criteria

Both tsunamis and storms events represent highly energetic physical processes capable of producing sedimentological shell beds (sensu Kidwell et al. [162]), although the resulting deposits display distinct sedimentological and taphonomic characteristics [25]. More recently, shell-bed concentrations have been widely used to infer their origins in both marine and lacustrine settings [12,15,19,24,25,37,43,45,76,257,292,301]. Studies on storm-generated shell beds (tempestites) have examined how taphonomic attributes such as fragmentation, orientation, and encrustation vary across different environments [259,286,294,302,303,304,305], while research on tsunami shell beds (tsunamites), has highlighted key taphonomic criteria related to extreme high-energy events (tsunami) [12,15,24,25,26,34,35,37,301].

From Oman, Donato et al. [12] present one the best examples applying bivalve-shell taphonomy to identification of tsunami deposits. They concluded that a key criterion characterizing such deposits is the mixture of allochthonous organisms derived from different environments (e.g., pelagic, shallow-water, and deep water, and even continental sources), coupled with a high degree of angular shell fragmentation. Similar signatures have been recognized in tsunamites from offshore settings associated with backwash flows. In this context, Puga-Bernabéu and Aguirre [19] compared the taphonomic features of two shell beds—one attributed to a tsunami and the other to a storm—and emphasized the importance of such comparisons for process discrimination. Comparable evidence has also been reported from coastal settings influenced by both run-up and backwash, including southwestern Spain [37] and northwestern Algeria [25].

Marine shell deposits linked to tsunamis are generally characterized by dense, thick, and laterally extensive beds [19]. The shells are often well to excellently preserved, frequently oriented obliquely or perpendicularly to bedding, and may be dominated by a single species [25,37]. Standardized 1 kg samples of fossiliferous sediments from tsunamigenic facies also yield consistently higher taxonomic richness and evenness than equivalent samples from tempestites [24], and these authors suggested that whenever fossils were present in tsunami deposits, this could be a decisive criterion to distinguish between tsunami and storm deposits. In addition, the low frequency of boring and encrustation combined at times with the presence of articulated bivalves in non-living position or disarticulated shells stacked like plates in dishwater (Figure 6), suggests that exceptionally powerful tsunami waves penetrated below the taphonomically active zone, eroding sediment masses that had previously buried the shells [13,15,20,25].

By contrast, tempestite shell beds usually consist of organisms derived from adjacent habitats, reflecting localized reworking by storm waves [19] (Figure 5). Their assemblages tend to exhibit lower taxonomic richness, higher rates of encrustation and bioerosion, and more ordered shell orientations. According to some authors [15,19,25,33,37] shells in tempestites are typically arranged subhorizontally, at angles of less than 30° to stratification, and concave-up stacking is common. Tsunamites, on the other hand, frequently display a wide range of depositional orientations, including chaotic fabrics, normal and reverse grading, as well as oblique to vertical alignments (>30–60°) (Figure 6). The abundance of large, heavy shells in molluscan assemblages has also been recognized as an additional criterion reflecting the extreme energy of tsunami waves in shallow coral reef lagoons [35]. Similarly, the chaotic and variable arrangement of concave shells in tsunamites, both from offshore [19] and onshore records [25,37], contrasts with the more systematic stacking observed in tempestites, further underscoring the diagnostic value of taphonomic features in distinguishing between these two types of extreme-wave deposits.

In addition to taphonomic criteria, comparison of microfossil taxonomic assemblages (e.g., ostracods, foraminifera, diatoms, and pollen) has also been successfully used to differentiate between storm and tsunami deposits [34,294,295,296,297,304,305]. Foraminiferal tests are frequently preserved in tsunami deposits and provide reliable evidence of sediment remobilization, often indicating the transport of seafloor material from depths greater than 100 m and from locations several kilometres offshore [221]. Similarly, diatom assemblages in tsunami deposits are often characterized by a chaotic mixture of freshwater and brackish–marine species, reflecting the entrainment of material from different environmental settings during transport [198] (Figure 5).

5.3. Geomorphological Criteria

In addition to sedimentological and taphonomic evidence, geomorphological indicators play a crucial role in identifying and distinguishing tsunami from storm deposits. Features such as boulders ridges, and landward sediment lobes provide insights into the dynamics of extreme-wave events and their capacity to reshape coastal landscapes [17,18,25,57,181,257]. These landforms provide crucial insights into the hydrodynamic energy, directionality, and inundation distance of extreme-wave events [16,20,28]. When preserved in the stratigraphic record, such geomorphological signatures serve as important proxies for identifying and reconstructing past tsunami events in both modern and ancient contexts. For Doukani et al. [25], the variations in facies and thickness in tsunamites are largely controlled by the pre-existing topography and the lithological characteristics of the affected areas (Figure 7). According to Nishimura and Miyaji [306] topographic undulations often control the thickness of sediment layers produced by a tsunami event. Such features have been documented both from the Pleistocene record [25] and recently, during the 2004 event along the west coast of Thailand [261], tsunami deposits tending to be thicker in channelized areas, whereas they are comparatively thinner over locally convex topographic highs. For Morton et al. [27] the lateral variability is not considered a key diagnostic criterion, as both tsunami and storm deposits may exhibit lateral variation in geometry. This variability is often strongly influenced by local costal and nearshore topography, including features such as river channels, canyons, embayments, and irrefular shorelines. Such topographic controls can modify flow paths, sediment transport, and deposition patterns, producing lateral differences in thickness, grain size, and bedding architecture that are not uniquely indicative of either tsunamis or storms. Therefore, while topography affects the distribution of deposits, lateral variability alone cannot reliably distinguish between these high-energy events.

Erosion zones produced by tsunami generally penetrate much farther inland and to much higher elevations above sea level than those generated by storms (Figure 5), owing to their greater wave amplitude, long period, and enhanced competence for sediment entrainment immediately beyond the shoreline [260,298,307]. By contrast, storm overwash deposits usually remain confined to the back-beach setting, often forming as extensions of the berm or beginning at an erosional scarp [299,308]. Broader erosion fields may develop under extreme storm conditions when wind stress substantially enhances flow [309]. Tsunamis also leave behind distinctive landform signatures, including deep scour channels, stripped soil layers, uprooted vegetation mats, and laterally continuous sand sheets or boulder ridges that may extend kilometers inland [25,28,34,43,263,277]. In similar depositional settings, the absence of marine sediments above and below tsunami layers is considered another key criterion for identifying tsunamites [25,37].

The transport of megaclasts onto coastal platforms or cliff tops is another hallmark rarely matched by storm activity [38,57,178,195]. Storms, in contrast, tend to generate localized geomorphic features such as overwash fans, washover terraces, and breaches through dune ridges, generally restricted to low-lying coastal barriers [299]. Taken together, blanketing of steep slopes by the sediments, often into elevations well above the reach of storms, the inland extent of erosion, the scale of sediment reworking, and the capacity to mobilize a metric megaclasts are key geomorphological criteria that help differentiate tsunami deposits from those generated by storms.

5.4. Geochemical Criteria

The distinctive geochemical signature has also been used to identify and interpret deposits related to tsunami events, e.g., [217,300]. The analysis of elemental geochemistry, often via X-ray fluorescence (XRF), further help the discrimination between tsunami and storm events. Tsunami deposits frequently exhibit enrichments in Strontium (Sr), Calcium (Ca), and Magnesium (Mg) related to the influx of shell fragments, carbonate sediments, and saline water [18,209,216,217] (Figure 8). Recent works analyzed the environmental DNA signatures for distinguishing between tsunami and storm deposits, e.g., [310]. Environmental DNA (eDNA) is a novel biogeochemical proxy for identifying tsunami deposits, with case studies from the Tohoku region (Japan) [300]. Here, the authors demonstrated that marine eDNA preserved in lacustrine sediments provided unequivocal evidence of seawater incursion during the 2011 Tohoku-oki tsunami, despite the inland setting of the deposits. Importantly, eDNA assemblages were also used to assess older events, including the 869 CE Jōgan tsunami and a prehistoric tsunami dated at 2400–2900 calibrated yr BP from the same region. In both cases, the eDNA composition of the event layers differed markedly from that of background sediments. Moreover, eDNA detected marine taxa in a thin muddy layer above the Jōgan deposit, which was visually indistinguishable from peat, suggesting that eDNA can reveal cryptic tsunami layers not recognized through sedimentological observation alone. This study highlights the potential of eDNA to complement classical geochemical indicators and to elucidate the origin of event deposits that are otherwise difficult to identify.

As mentioned above, it is often difficult to determine whether tsunami succession units are the result of a single event or of multiple events that occurred either close together or separated in time. Moreira et al. [218] demonstrated that high-resolution XRF core scanning, combined with detailed grain-size and image analyses, can help disentangle the complexity of tsunami deposition. Their study of the 1755 Lisbon tsunami revealed up to four distinct depositional pulses, interpreted as successive inundation and backwash phases, thereby illustrating how geochemical signatures and textural variability can be used to refine event stratigraphy and reconstruct depositional phases.

6. Conclusions

This review covers the global and temporal distribution of tsunami deposits from the Precambrian to recent times, underscoring their value as persistent recorders of catastrophic events throughout Earth’s history. The geological record, from the earliest tsunamites triggered by bolide impacts in the Precambrian to those associated with major mass extinctions and the well-calibrated events of the Holocene, provides an unparalleled long-term perspective. It reveals that tsunamis are a recurring geological phenomenon, generated by a variety of triggers including earthquakes, submarine landslides, volcanic flank collapses, volcanic eruptions, extraterrestrial impacts, and climate-driven processes. The preservation of these event-deposits across the geological times highlights the importance of tsunamites not only for understanding past dynamics of the Earth system, but also for assessing future coastal hazards.

Whereas the Pleistocene and Holocene records are relatively robust, older successions from the Precambrian to the Cenozoic remain underexplored, presenting a significant knowledge gap. Future studies should focus on applying advanced geochemical, taphonomic, sedimentological, geomorphological and, whenever possible, quantitative palaeontological approaches to these ancient deposits, while continuing to use modern analogues for calibration.

A principal finding of this synthesis is the critical importance and necessity of a multi-proxy methodology for the accurate identification of tsunami deposits and their discrimination from tempestites. As detailed across the sedimentological, taphonomic, geomorphological, geochemical, and palaeontological criteria, no single feature is unequivocally diagnostic. The most confident interpretations arise from the convergence of multiple lines of evidence: (i) the presence of erosive bases and rip-up clasts coupled with a chaotic grain-size distribution; (ii) taphonomic signatures such as mixed ecological assemblages and variable shell orientations; (iii) geomorphological indicators like far-inland penetration and megaclast emplacement; (iv) geochemical tracers of marine incursion, including ions, elemental ratios, and now, revolutionary environmental DNA; (v) and quantitative palaeontological methods. This integrated approach is paramount for avoiding misinterpretation in the geological record and for building reliable paleotsunami databases.